I simulate a cyclone separator using FLUENT6.1, and I get a higher swirling flow than my experiments. The simulated tangential velocity is larger than experimental data, and the recirculation zone appears, while the experiments don't have that; the simulated pressure drop is also larger than experimental results. When I want to reduce some degree of swirl by adding 0.0002m roughness, the Rankine vortex velocity profiles was destroyed again. My compution method:RSM, SIMPLE, QUICK. How to sovle above question? please give me some advice! thank you very much

LCW

Anyone do the simulation of cyclone separator? welcome to present some experiences.